Accurate measurements of serum cholesterol are vital for the proper screening of patients for high and borderline-high cholesterol as part of the National Cholesterol Education Program. Some of the enzymatic methods (such as the aca from DuPont) for cholesterol show a bias compared with the Reference Method. Further, some of the materials used to check the accuracy of methods in the field (such as the CAP survey materials and the SRM 909 from NIS) show discrepancies when cholesterol is determined with different methods. Lyophilization of the reference material and calibrators is a key factor in the biases and discrepancies seen. The effect of lyophilization on the evaluation for linearity for HDL- cholesterol was investigated.